Machine Learning (ML)

Machine Learning (ML) is a subset of Artificial Intelligence that allows computers to “learn” from data. Ordinarily, in programming, we provide data and the expected output, and the machine does the work. In ML, it's a bit different; its goal is generalization, that is, training a model so it can use data it has never seen before and make a fairly accurate description. For ML to work, we need large collections of data. This data is then prepared/cleaned so the machine can read it - like turning text into binary numbers. Next comes the training, the algorithm analyzes the data in order to find patterns, for example, like in spam detection. The machine “learns” by being trained with a large dataset, and at the end, we can use unknown data to get results. ML can be supervised, unsupervised, and refined. Real-world examples of machine learning are the suggestions you see on platforms like YouTube or Netflix, the facial recognition feature on your phone, and language translation tools. In a nutshell, machines are much faster than humans in recognizing repeating patterns, and ML can be very useful, or very harmful, depending on how it’s used.

1. Supervised learning trains a model on labelled data, where every example comes with the correct answer. Used for spam filtering and predicting house prices.
2. Unsupervised learning finds patterns in unlabelled data without being told what to look for. Used for customer clustering and fraud detection.
3. Reinforcement learning teaches an agent through trial and error, using rewards and penalties. Used in game-playing AI and robotics.
4. Semi-supervised learning combines a small amount of labelled data with a large amount of unlabelled data. Useful when labelling is expensive.
5. Self-supervised learning has the model create its own labels from raw data, such as predicting the next word in a sentence. This is how large language models are pre-trained.